TabNine
TabNine, An autocomplete feature is something that helps in replying to tons of emails or making quick responses. But what if this feature could also help developers be more productive?
Deep Tabnine is such an autocompleter tool that uses the power of deep learning. This tool works on the simple principle of text prediction. Deep Tabnine uses GPT-2 which is a natural language processing model to generate relevant coding suggestions and tokens based on the tokens you have typed in.
Deep Tabnine works with existing IDEs and code editors. Currently, it supports 22 languages in total including Swift, Objective-C, JavaScript, PHP, Java, Rust, Go and C#.
Jacob Jackson, the creator of Tabnine, has said that Deep Tabnine can give you a distribution of tokens that you are going to see next if you have a sequence of code.
It automatically compiles and includes the probability of different predictions based on the past usage and habit of the programmer.
Why TabNine?
- TabNine does not require any configuration in order to work.
- It works for all programming languages.
- TabNine does not require any external software.
Using Deep TabNine
It is trained on about two million files on GitHub this helps the model to learn some complex behaviors in dynamically typed languages.
Deep Tabnine can infer function names, its parameters and return types from a document that is written in natural language.
It requires a lot of computing power: running the model on a laptop would not deliver the low latency that TabNine’s users have come to expect. So we are offering a service that will allow you to use TabNine’s servers for GPU-accelerated autocompletion. It’s called TabNine Cloud.
TabNine Cloud is currently in beta, and scaling it up presents some unique challenges since queries are computationally yet they must be fulfilled with low latency.
If you like to discuss it in detail. Please, contact us on Email or on social media(twitter/facebook)
u